Preferences-based choice prediction in evolutionary multi-objective optimization

Manish Aggarwal; Justin Heinermann; Stefan Oehmcke; Oliver Kramer

doi:10.1007/978-3-319-55849-3_46

Preferences-based choice prediction in evolutionary multi-objective optimization

Manish Aggarwal, Justin Heinermann, Stefan Oehmcke, Oliver Kramer^*

^*Corresponding author for this work

2 Citations (Scopus)

Abstract

Evolutionary multi-objective algorithms (EMOAs) of the type of NSGA-2 approximate the Pareto-front, after which a decisionmaker (DM) is confounded with the primary task of selecting the best solution amongst all the equally good solutions on the Pareto-front. In this paper, we complement the popular NSGA-2 EMOA by posteriori identifying a DM’s best solution among the candidate solutions on the Pareto-front, generated through NSGA-2. To this end, we employ a preference-based learning approach to learn an abstract ideal reference point of the DM on the multi-objective space, which reflects the compromises the DM makes against a set of conflicting objectives. The solution that is closest to this reference-point is then predicted as the DM’s best solution. The pairwise comparisons of the candidate solutions provides the training information for our learning model. The experimental results on ZDT1 dataset shows that the proposed approach is not only intuitive, but also easy to apply, and robust to inconsistencies in the DM’s preference statements.

Original language	English
Title of host publication	Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings
Editors	J.Ignacio Hidalgo, Carlos Cotta, Ting Hu, Alberto Tonda, Paolo Burrelli, Matt Coler, Giovanni Iacca, Michael Kampouridis, Antonio M. Mora Garcia, Giovanni Squillero, Anthony Brabazon, Evert Haasdijk, Jacqueline Heinerman, Fabio D Andreagiovanni, Jaume Bacardit, Trung Thanh Nguyen, Sara Silva, Ernesto Tarantino, Anna I. Esparcia-Alcazar, Gerd Ascheid, Kyrre Glette, Stefano Cagnoni, Paul Kaufmann, Francisco Fernandez de Vega, Michalis Mavrovouniotis, Mengjie Zhang, Federico Divina, Kevin Sim, Neil Urquhart, Robert Schaefer
Number of pages	10
Publisher	Springer Verlag,
Publication date	1 Jan 2017
Pages	715-724
ISBN (Print)	9783319558486
DOIs	https://doi.org/10.1007/978-3-319-55849-3_46
Publication status	Published - 1 Jan 2017
Event	20th European Conference on the Applications of Evolutionary Computation, EvoApplications 2017 - Amsterdam, Netherlands Duration: 19 Apr 2017 → 21 Apr 2017

Conference

Conference	20th European Conference on the Applications of Evolutionary Computation, EvoApplications 2017
Country/Territory	Netherlands
City	Amsterdam
Period	19/04/2017 → 21/04/2017

Series	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10199 LNCS
ISSN	0302-9743

Keywords

Multi-objective optimization
NSGA-2
Preference-based learning
Solution selection

Access to Document

10.1007/978-3-319-55849-3_46

Cite this

Aggarwal, M., Heinermann, J., Oehmcke, S., & Kramer, O. (2017). Preferences-based choice prediction in evolutionary multi-objective optimization. In J. I. Hidalgo, C. Cotta, T. Hu, A. Tonda, P. Burrelli, M. Coler, G. Iacca, M. Kampouridis, A. M. Mora Garcia, G. Squillero, A. Brabazon, E. Haasdijk, J. Heinerman, F. D Andreagiovanni, J. Bacardit, T. T. Nguyen, S. Silva, E. Tarantino, A. I. Esparcia-Alcazar, G. Ascheid, K. Glette, S. Cagnoni, P. Kaufmann, F. F. de Vega, M. Mavrovouniotis, M. Zhang, F. Divina, K. Sim, N. Urquhart, ... R. Schaefer (Eds.), Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings (pp. 715-724). Springer Verlag,. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Vol. 10199 LNCS https://doi.org/10.1007/978-3-319-55849-3_46

Preferences-based choice prediction in evolutionary multi-objective optimization. / Aggarwal, Manish; Heinermann, Justin; Oehmcke, Stefan et al.

Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings. ed. / J.Ignacio Hidalgo; Carlos Cotta; Ting Hu; Alberto Tonda; Paolo Burrelli; Matt Coler; Giovanni Iacca; Michael Kampouridis; Antonio M. Mora Garcia; Giovanni Squillero; Anthony Brabazon; Evert Haasdijk; Jacqueline Heinerman; Fabio D Andreagiovanni; Jaume Bacardit; Trung Thanh Nguyen; Sara Silva; Ernesto Tarantino; Anna I. Esparcia-Alcazar; Gerd Ascheid; Kyrre Glette; Stefano Cagnoni; Paul Kaufmann; Francisco Fernandez de Vega; Michalis Mavrovouniotis; Mengjie Zhang; Federico Divina; Kevin Sim; Neil Urquhart; Robert Schaefer. Springer Verlag, 2017. p. 715-724 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 10199 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Aggarwal, M, Heinermann, J, Oehmcke, S & Kramer, O 2017, Preferences-based choice prediction in evolutionary multi-objective optimization. in JI Hidalgo, C Cotta, T Hu, A Tonda, P Burrelli, M Coler, G Iacca, M Kampouridis, AM Mora Garcia, G Squillero, A Brabazon, E Haasdijk, J Heinerman, F D Andreagiovanni, J Bacardit, TT Nguyen, S Silva, E Tarantino, AI Esparcia-Alcazar, G Ascheid, K Glette, S Cagnoni, P Kaufmann, FF de Vega, M Mavrovouniotis, M Zhang, F Divina, K Sim, N Urquhart & R Schaefer (eds), Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings. Springer Verlag, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10199 LNCS, pp. 715-724, 20th European Conference on the Applications of Evolutionary Computation, EvoApplications 2017, Amsterdam, Netherlands, 19/04/2017. https://doi.org/10.1007/978-3-319-55849-3_46

Aggarwal M, Heinermann J, Oehmcke S, Kramer O. Preferences-based choice prediction in evolutionary multi-objective optimization. In Hidalgo JI, Cotta C, Hu T, Tonda A, Burrelli P, Coler M, Iacca G, Kampouridis M, Mora Garcia AM, Squillero G, Brabazon A, Haasdijk E, Heinerman J, D Andreagiovanni F, Bacardit J, Nguyen TT, Silva S, Tarantino E, Esparcia-Alcazar AI, Ascheid G, Glette K, Cagnoni S, Kaufmann P, de Vega FF, Mavrovouniotis M, Zhang M, Divina F, Sim K, Urquhart N, Schaefer R, editors, Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings. Springer Verlag,. 2017. p. 715-724. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 10199 LNCS). doi: 10.1007/978-3-319-55849-3_46

Aggarwal, Manish ; Heinermann, Justin ; Oehmcke, Stefan et al. / Preferences-based choice prediction in evolutionary multi-objective optimization. Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings. editor / J.Ignacio Hidalgo ; Carlos Cotta ; Ting Hu ; Alberto Tonda ; Paolo Burrelli ; Matt Coler ; Giovanni Iacca ; Michael Kampouridis ; Antonio M. Mora Garcia ; Giovanni Squillero ; Anthony Brabazon ; Evert Haasdijk ; Jacqueline Heinerman ; Fabio D Andreagiovanni ; Jaume Bacardit ; Trung Thanh Nguyen ; Sara Silva ; Ernesto Tarantino ; Anna I. Esparcia-Alcazar ; Gerd Ascheid ; Kyrre Glette ; Stefano Cagnoni ; Paul Kaufmann ; Francisco Fernandez de Vega ; Michalis Mavrovouniotis ; Mengjie Zhang ; Federico Divina ; Kevin Sim ; Neil Urquhart ; Robert Schaefer. Springer Verlag, 2017. pp. 715-724 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 10199 LNCS).

@inproceedings{ec1d2766b3044033b3b1efd5b065bf9c,

title = "Preferences-based choice prediction in evolutionary multi-objective optimization",

abstract = "Evolutionary multi-objective algorithms (EMOAs) of the type of NSGA-2 approximate the Pareto-front, after which a decisionmaker (DM) is confounded with the primary task of selecting the best solution amongst all the equally good solutions on the Pareto-front. In this paper, we complement the popular NSGA-2 EMOA by posteriori identifying a DM{\textquoteright}s best solution among the candidate solutions on the Pareto-front, generated through NSGA-2. To this end, we employ a preference-based learning approach to learn an abstract ideal reference point of the DM on the multi-objective space, which reflects the compromises the DM makes against a set of conflicting objectives. The solution that is closest to this reference-point is then predicted as the DM{\textquoteright}s best solution. The pairwise comparisons of the candidate solutions provides the training information for our learning model. The experimental results on ZDT1 dataset shows that the proposed approach is not only intuitive, but also easy to apply, and robust to inconsistencies in the DM{\textquoteright}s preference statements.",

keywords = "Multi-objective optimization, NSGA-2, Preference-based learning, Solution selection",

author = "Manish Aggarwal and Justin Heinermann and Stefan Oehmcke and Oliver Kramer",

year = "2017",

month = jan,

day = "1",

doi = "10.1007/978-3-319-55849-3_46",

language = "English",

isbn = "9783319558486",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag,",

pages = "715--724",

editor = "J.Ignacio Hidalgo and Carlos Cotta and Ting Hu and Alberto Tonda and Paolo Burrelli and Matt Coler and Giovanni Iacca and Michael Kampouridis and {Mora Garcia}, {Antonio M.} and Giovanni Squillero and Anthony Brabazon and Evert Haasdijk and Jacqueline Heinerman and {D Andreagiovanni}, Fabio and Jaume Bacardit and Nguyen, {Trung Thanh} and Sara Silva and Ernesto Tarantino and Esparcia-Alcazar, {Anna I.} and Gerd Ascheid and Kyrre Glette and Stefano Cagnoni and Paul Kaufmann and {de Vega}, {Francisco Fernandez} and Michalis Mavrovouniotis and Mengjie Zhang and Federico Divina and Kevin Sim and Neil Urquhart and Robert Schaefer",

booktitle = "Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings",

note = "20th European Conference on the Applications of Evolutionary Computation, EvoApplications 2017 ; Conference date: 19-04-2017 Through 21-04-2017",

}

TY - GEN

T1 - Preferences-based choice prediction in evolutionary multi-objective optimization

AU - Aggarwal, Manish

AU - Heinermann, Justin

AU - Oehmcke, Stefan

AU - Kramer, Oliver

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Evolutionary multi-objective algorithms (EMOAs) of the type of NSGA-2 approximate the Pareto-front, after which a decisionmaker (DM) is confounded with the primary task of selecting the best solution amongst all the equally good solutions on the Pareto-front. In this paper, we complement the popular NSGA-2 EMOA by posteriori identifying a DM’s best solution among the candidate solutions on the Pareto-front, generated through NSGA-2. To this end, we employ a preference-based learning approach to learn an abstract ideal reference point of the DM on the multi-objective space, which reflects the compromises the DM makes against a set of conflicting objectives. The solution that is closest to this reference-point is then predicted as the DM’s best solution. The pairwise comparisons of the candidate solutions provides the training information for our learning model. The experimental results on ZDT1 dataset shows that the proposed approach is not only intuitive, but also easy to apply, and robust to inconsistencies in the DM’s preference statements.

AB - Evolutionary multi-objective algorithms (EMOAs) of the type of NSGA-2 approximate the Pareto-front, after which a decisionmaker (DM) is confounded with the primary task of selecting the best solution amongst all the equally good solutions on the Pareto-front. In this paper, we complement the popular NSGA-2 EMOA by posteriori identifying a DM’s best solution among the candidate solutions on the Pareto-front, generated through NSGA-2. To this end, we employ a preference-based learning approach to learn an abstract ideal reference point of the DM on the multi-objective space, which reflects the compromises the DM makes against a set of conflicting objectives. The solution that is closest to this reference-point is then predicted as the DM’s best solution. The pairwise comparisons of the candidate solutions provides the training information for our learning model. The experimental results on ZDT1 dataset shows that the proposed approach is not only intuitive, but also easy to apply, and robust to inconsistencies in the DM’s preference statements.

KW - Multi-objective optimization

KW - NSGA-2

KW - Preference-based learning

KW - Solution selection

UR - http://www.scopus.com/inward/record.url?scp=85017558458&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-55849-3_46

DO - 10.1007/978-3-319-55849-3_46

M3 - Article in proceedings

AN - SCOPUS:85017558458

SN - 9783319558486

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 715

EP - 724

BT - Applications of Evolutionary Computation - 20th European Conference, EvoApplications 2017, Proceedings

A2 - Hidalgo, J.Ignacio

A2 - Cotta, Carlos

A2 - Hu, Ting

A2 - Tonda, Alberto

A2 - Burrelli, Paolo

A2 - Coler, Matt

A2 - Iacca, Giovanni

A2 - Kampouridis, Michael

A2 - Mora Garcia, Antonio M.

A2 - Squillero, Giovanni

A2 - Brabazon, Anthony

A2 - Haasdijk, Evert

A2 - Heinerman, Jacqueline

A2 - D Andreagiovanni, Fabio

A2 - Bacardit, Jaume

A2 - Nguyen, Trung Thanh

A2 - Silva, Sara

A2 - Tarantino, Ernesto

A2 - Esparcia-Alcazar, Anna I.

A2 - Ascheid, Gerd

A2 - Glette, Kyrre

A2 - Cagnoni, Stefano

A2 - Kaufmann, Paul

A2 - de Vega, Francisco Fernandez

A2 - Mavrovouniotis, Michalis

A2 - Zhang, Mengjie

A2 - Divina, Federico

A2 - Sim, Kevin

A2 - Urquhart, Neil

A2 - Schaefer, Robert

PB - Springer Verlag,

T2 - 20th European Conference on the Applications of Evolutionary Computation, EvoApplications 2017

Y2 - 19 April 2017 through 21 April 2017

ER -

Preferences-based choice prediction in evolutionary multi-objective optimization

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this